Skip to contentSkip to conversation

net: pktgen: fix race between pktgen_thread_worker() and kthread_stop()

PATCH In review Oleg Nesterov 3 participants 5 messages Jul 8, 2015 – Jul 10, 2015

Subsystems (2)
Networking Network drivers
lore
Share & tools
lore

Oleg Nesterov

author
#0 lore raw

cover net: pktgen: fix race between pktgen_thread_worker() and kthread_stop()

Hello,

I am not familiar with this code and I have no idea how to test these changes, so 2/2 comes as a separate change. 1/2 looks like the obvious bugfix, and probably candidate for -stable.

Oleg.

 net/core/pktgen.c |    9 ++-------
 1 files changed, 2 insertions(+), 7 deletions(-)
author
#1 lore raw

PATCH 1/2 net: pktgen: fix race between pktgen_thread_worker() and kthread_stop()

pktgen_thread_worker() is obviously racy, kthread_stop() can come between the kthread_should_stop() check and set_current_state().

Signed-off-by Oleg Nesterov <[email protected]>
Reported-by Jan Stancek <[email protected]>
Reported-by Marcelo Leitner <[email protected]>
---
 net/core/pktgen.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)
net/core/pktgen.c +3 -1
index 508155b..043ea18 100644
@@ -3490,8 +3490,10 @@ static int pktgen_thread_worker(void *arg)
 	pktgen_rem_thread(t);
 
 	/* Wait for kthread_stop */
-	while (!kthread_should_stop()) {
+	for (;;) {
 		set_current_state(TASK_INTERRUPTIBLE);
+		if (kthread_should_stop())
+			break;
 		schedule();
 	}
 	__set_current_state(TASK_RUNNING);
Signature · 1 line
1.5.5.1
author
#2 lore raw

PATCH 2/2 net: pktgen: kill the "Wait for kthread_stop" code in pktgen_thread_worker()

In reply to Oleg Nesterov

pktgen_thread_worker() doesn't need to wait for kthread_stop(), it can simply exit. Just pktgen_create_thread() and pg_net_exit() should do get_task_struct()/put_task_struct(). kthread_stop(dead_thread) is fine.

Signed-off-by Oleg Nesterov <[email protected]>
---
 net/core/pktgen.c |   11 ++---------
 1 files changed, 2 insertions(+), 9 deletions(-)
net/core/pktgen.c +2 -9
index 043ea18..8e0181a 100644
@@ -3489,15 +3489,6 @@ static int pktgen_thread_worker(void *arg)
 	pr_debug("%s removing thread\n", t->tsk->comm);
 	pktgen_rem_thread(t);
 
-	/* Wait for kthread_stop */
-	for (;;) {
-		set_current_state(TASK_INTERRUPTIBLE);
-		if (kthread_should_stop())
-			break;
-		schedule();
-	}
-	__set_current_state(TASK_RUNNING);
-
 	return 0;
 }
 

@@ -3689,6 +3680,7 @@ static int __net_init pktgen_create_thread(int cpu, struct pktgen_net *pn)
 	}
 
 	t->net = pn;
+	get_task_struct(p);
 	wake_up_process(p);
 	wait_for_completion(&t->start_done);
 

@@ -3811,6 +3803,7 @@ static void __net_exit pg_net_exit(struct net *net)
 		t = list_entry(q, struct pktgen_thread, th_list);
 		list_del(&t->th_list);
 		kthread_stop(t->tsk);
+		put_task_struct(t->tsk);
 		kfree(t);
 	}
Signature · 1 line
1.5.5.1
maintainer
#3 lore raw
In reply to Oleg Nesterov

From: Oleg Nesterov <[email protected]>
Date: Wed, 8 Jul 2015 21:41:54 +0200

Quoted 3 lines

I am not familiar with this code and I have no idea how to test these changes, so 2/2 comes as a separate change. 1/2 looks like the obvious bugfix, and probably candidate for -stable.

These look fine to me, applied and patch #1 queued up for -stable.

Thanks.

#4 lore raw
In reply to Oleg Nesterov
Quoted 11 lines · Oleg Nesterov

Hello,

I am not familiar with this code and I have no idea how to test these changes, so 2/2 comes as a separate change. 1/2 looks like the obvious bugfix, and probably candidate for -stable.

Oleg.

 net/core/pktgen.c |    9 ++-------
 1 files changed, 2 insertions(+), 7 deletions(-)

Both patches tested here, works for me.

Thanks,
Marcelo